There is a gap in in the availability of predictive biomarkers that are necessary to personalize treatment for patients with GBM. A challenge has been to separate prognostic makers (those that predict the natural history of the disease) from predictive (those that predict response to a particular treatment). l/Vfe hypothesize that effective biomarkers of treatment response must be decoupled from prognostic influences. Furthermore, to maximize clinical utility, these biomarkers must be developed as high-throughput, validated diagnostics applicable to small quantities of formalin-fixed, paraffin-embedded (FFPE) tissue. The specific aims of Project 3 are: 1) Develop a novel clinical diagnostic to classify patients suitable for FFPE tissue and incorporating the most robust molecular prognostic markers. 2) Develop a predictive marker of response to anti-angiogenic therapy. 3) Develop a predictive marker of response to immunomodulation. In the initial MDACC Brain SPORE funding, we developed a 9-gene FFPE biomarker that distinguished favorable from unfavorable patient prognosis. The assay was validated, commercialized, and incorporated as a prospective stratification factor in a phase III trial evaluating the anti-angiogenesis inhibitor bevacizumab [Radiation Therapy Oncology Group (RTOG) 0825]. To further improve survival discrimination, we expanded the 9-gene signature to a more robust 19-gene. In addition, we identified an additional epigenetic prognostic signature, the glioma CpG-island methlylator phenotype (G-CIMP). Using >300 cases from M.D. Anderson Cancer Center (MDACC) we developed a unified four risk group system called the molecular-clinical, prognosticator (MCP) that combines clinicopathologic factors with gene expression, genetic, and epigenetic factors. The analytic approach in constructing the MCP and its improved prognostic classification will allow us to better develop predictive makers independent of prognostic influences. Based on our success in the initial SPORE funding period and leveraging our expertise in biomarker development and our close involvement with the RTOG we know propose renewal of this project.